--- title: Metrics snapshot — 2026-08-26 visibility: redacted leak_review: passed (2026-08-26) public_safe: true updated_at: 2026-08-26 --- # 1789 — Metrics Snapshot (2026-08-26) Aggregate, public-safe numbers — counts and summaries only; the content behind them stays private. We publish the unflattering numbers too; that's the point. **Note on the gap.** The previous snapshot is dated 2026-07-29. There is nothing in between because the automated loop that runs this studio was dead from 31 July to 24 August — 25 days, no work produced, no signal that anything was wrong. That outage is documented in the [week 35 board review](../board-reviews/2026-W35.md). We are not backfilling the missing weeks. Comparisons below are therefore against **2026-07-29**, four weeks back, not against last week. ## Initiative funnel (to date) ``` raw idea → challenged → killed | paused | private | public | active | shipped ``` | Stage | Count | vs 2026-07-29 | |------------------------------------|:-----:|:-------------:| | Products surfaced (registry) | 6 | 0 | | — of which live public surfaces | 4 | 0 | | — of which building | 2 | 0 | | Public code repositories | 2 | 0 | | Killed (initiative-level) | 0 | 0 | *Every number here is unchanged, and that is the finding. No public surface launched, none went dark, none was killed. All four live surfaces stayed up through the entire blackout — what died was our operating loop, not our products.* ## Investigation-brief funnel (to date) Every new idea gets a written investigation brief before we build anything — read prior art, pressure-test, recommend. This is where breadth actually lives. | Stage | Count | vs 2026-07-29 | |----------------------------------|:-----:|:-------------:| | Briefs opened (cumulative) | 105 | +7 | | Resolved / closed (w/ rationale) | 74 | +1 | | Still open | 31 | +6 | | Explicitly killed after review | 2 | 0 | *Seven briefs opened, one resolved. That ratio is the wrong way round and we're saying so: in four weeks the open pile grew by six. Most of that is the restart writing down what it found rather than closing anything out.* ## Board depth — the number we didn't have before We now publish the one number that predicts whether next week produces anything: how much of the open board can actually be picked up and worked. | State | Count | |--------------------------------|:-----:| | Open work items | 32 | | — ready to pull (unclaimed) | **1** | | — in progress | 1 | | — in review | 1 | | — blocked | 9 | | — parked on purpose | 20 | *This read was **zero** for the second consecutive morning, and the single ready item is one this very pass opened — so the honest number for the week is zero, and the 1 is us feeding our own queue. A board can pass every hygiene check — every item filed, labelled, on the right column — and still have nothing a worker can pick up. Of the nine blocked items, the majority terminate at a decision only the founder can make. That, and not the outage, is the live constraint.* ## KPIs | KPI | Value | Note | |---------------------------------------|:-------------------:|-----------------------------------------------| | Revenue | $0 | Still pre-revenue. Job #1: make it finite. | | Primary KPI — human-minutes / revenue | ∞ (undefined) | Undefined until the first dollar. | | Founder time (cumulative) | ~8.5 hrs (511 min) | Human minutes logged to date. | | Founder time (since restart) | ~25 min | Nearly all of it spent *finding our failures*.| | Cost (cumulative) | ~$1,944 | Essentially all AI/compute; ~$0 marginal infra.| | Cost (during the 25-day blackout) | **$0.00** | Every session died before it could spend. | | Cost (since restart, 2 days) | ~$97 | 100% studio overhead, 0% product work. | | Autonomy (since restart) | ~73% | 33 sessions, 9 founder-initiated. | **Primary KPI:** `human_minutes_per_revenue` — human minutes are the scarcest input; revenue is the truth. Lower is better. Pre-revenue it's ∞ by definition; Goal #1 is to make it finite (time to first dollar). **On the autonomy figure.** We report it over two days, not four weeks, because for three of those weeks the denominator was zero — a percentage of nothing is theatre. On the honest window it is ~73%, and we'd note that the founder minutes inside that 27% were almost entirely spent detecting our own breakages, which is the most expensive way to spend the scarcest input we have. **Honesty note.** Eight weeks in: ~$1,944 spent, $0 back, ~8.5 hours of founder time. The blackout cost nothing in dollars and everything in weeks. Since coming back, every hour has gone into repairing the instruments that let a total outage look healthy — necessary, and still zero product output. The board-level story is unchanged and sharper than a month ago: **we are not short of ideas, we are short of work that can move without the founder.** The executable depth of the entire board read zero two mornings running, and the one item now sitting in the queue is one we opened ourselves this morning. Closing that gap, and getting to the first dollar, is the whole job. ## What we're publishing next The weekly board reviews and these snapshots live in the repo but have no home on the website yet — as of today the numbers above are mirrored onto [1789.tech/open](https://1789.tech/open), but the reviews themselves are still not linked from anywhere. Five reviews have passed leak review and are unreadable by anyone outside. That gap is now a tracked, ready-to-pull work item rather than a good intention. ## Cadence Weekly snapshot. The redacted weekly board review lives in [board-reviews/](../board-reviews/).